It is time to use AI at scale for real!
Are you an asset owner or have a customer who wishes to understand the advantages of artificial intelligence for energy and utility companies? Sterblue offers you a free AI test which will allow you to see the results on your own inspection images. The entire process is expected to be completed in roughly 2 weeks.
We’ve launched this initiative early October 2020 and 5 major utility companies in the US and Europe have already jumped on board. If you want to join them, please submit this form.
Why should you run Sterblue's AI test?
Showcase AI capabilities inside your organization:
Over the past 4 years, Sterblue has developed its unique AI software system called Curiosity, which is capable of detecting defects and inventory on transmission and distribution grid, while accounting with infrastructure heterogeneity worldwide. Examples of common detections are corrosion spots, broken insulators, woodpecker holes, inclined poles, insulators, transformers, wooden poles, and so on.
Extract key insights for asset health monitoring:
Sure AI can be an impressive tool, which helps improve the productivity of industry experts. But what can you actually do with the thousands of detections it outputs? As said, AI is just a tool, which can be integrated into your inspection workflow and help you optimize it, by enriching GIS, building health indexes, assessing corrosion or planning maintenance.
Build a business case on the added value of AI:
Demonstrate inside your organization with your data the advantages of adding AI to the inspection workflow. The benefits will be represented in short or long term KPIs, such as time savings, increased reliability, increased productivity and others, depending on the customer’s business objectives.
“We were impressed by the outputs of Sterblue AI test! Sterblue was able to properly detect equipment and defects on our distribution & transmission grid. On top of this, the implementation of the test was pretty smooth, with a business case report being delivered in the end, stating insights we could easily share internally.”
What does Sterblue need to get the test running?
- From 1,000 to 2,000 homogeneous images with the defects or equipment to be detected taken from the ground, helicopter or drone.
- Ideally, your quality nomenclature and, when existing, a CSV report matching photos with defect/equipment type & severity level.
What will be the process during approximately 2 weeks?
- Kickoff meeting to align expectations with the customer (30-45mins)
- Sterblue analyzes the set of images provided and assess what is most relevant to demonstrate from an automation perspective.
- Sterblue trains and fine tunes its dedicated AI on 80% of the set of images provided, to make sure we take into account the specifications of your infrastructure and detections.
- Run Sterblue AI on the remaining 20% of the image set to demonstrate the actual results. If you prefer, you can also provide us with an extra 200-300 images and we will run the test live, during the delivery meeting.
- Delivery meeting where you will get to see the results of the AI directly on Sterblue Cloud platform, as well our report capabilities. Here you will have access to different KPIs which can be customized to your business objectives as well as exported to other internal software systems.
For a more detailed overview of how the Sterblue Team handles an AI test and the different challenges and learnings, you should definitely read our newest white paper on how in only a few days we prepared, trained and tested our AI on the defects and specifications required by a major TSO in Europe.
Ready to get started? Submit this form and we will reach out to you!